AbstractThe aim of this study is to optimize rolling forces in a multi-pass hot rolling process using multiple linear regression (MLR) and the artificial bee colony (ABC) algorithm. In multi-pass rolling, a homogeneous force distribution enhances material quality by creating a uniform internal structure and improving mechanical properties. Imbalanced forces can cause premature wear of equipment, so a homogeneous distribution reduces wear and extends machine lifespan. It also allows for better process control, enhancing stability and preventing undesired deformations, which ultimately leads to lower energy consumption and increased production efficiency. This study focused on a five-pass roughing process in a bar rolling mill. Initial and final material dimensions were defined, and the spread and rolling forces for each pass were calculated using MLR, based on experiments conducted at a constant temperature. The ABC algorithm was subsequently employed to optimize the calculated values. The results were validated through experimental testing. The ABC algorithm achieved a maximum error of 1.6% in spread and 20% in rolling force. In comparison, the Finite Element Method (FEM) yielded errors of 1.1% in spread and 17.2% in rolling force. These findings demonstrate that optimization using MLR and the ABC algorithm can be performed both quickly and with high accuracy. Furthermore, the performance of the ABC algorithm in finding optimal solutions under different parameter configurations was investigated. The results indicate that as the number of iterations, food sources, and limit values increase, the probability of reaching optimal solutions also improves, suggesting that each parameter contributes differently to the optimization process. Optimizing the rolling forces resulted in a 69.4% reduction in weight differences and a 66.6% decrease in standard deviation, thereby improving dimensional uniformity and product quality. Additionally, the implementation of the new design led to a significant 10.4% reduction in energy consumption.
:37
:The International Journal of Advanced Manufacturing Technology